Speech-to-Text Translation on libri-trans
Metric: Case-sensitive tokenized BLEU (higher is better)
Metric: Case-sensitive tokenized BLEU (higher is better)
| # | Model↕ | Case-sensitive tokenized BLEU▼ | Extra Data | Paper | Date↕ | Code |
|---|---|---|---|---|---|---|
| 1 | Transformer + ASR Pretrain + SpecAug | 17.8 | No | NeurST: Neural Speech Translation Toolkit | 2020-12-18 | Code |
| 2 | Transformer + ASR Pretrain | 16.9 | No | NeurST: Neural Speech Translation Toolkit | 2020-12-18 | Code |